Paul GRIMWOOD

LONDON FIRE BRIGADE retd
The concept for using redundant steel shipping containers to teach firefighters how a compartment fire is likely to develop and behave under variable ventilation parameters was introduced by Swedish fire officers Mats Rosander and Anders Lauren during the early 1980s. The containers were designed to simulate, as realistically as possible, the formation and transport of fire gases within a compartment whilst demonstrating a range of phenomena related to 'flashover', backdraft and other forms of fire gas ignitions. These specific fireground hazards were increasingly becoming linked with firefighter deaths and quite often this was because they failed to understand the basic principles of fire development and fire behavior within the confined state of a structure. The simulators also provided an opportunity to practice counter tactics for dealing with fire gases accumulating and igniting in the overhead. The introduction of 3D water-fog applications and tactical venting actions were central to the 'safe-person' concepts and methods of operational risk assessment being developed in the UK and Sweden.

The methods and tactics used in the simulators were to vastly improve firefighter safety at fires over the following years and several fire authorities in Sweden, Finland, UK, Germany, Australia, Spain and the USA were the first to adopt the 'new-wave' approaches in dealing with confined and under-ventilated fires as well as blazing 'reservoirs' of fire gases existing in stair-shafts, voids and compartments. Whilst gas-fired training facilities offered an environmentally friendly alternative to chip-board linings and 'real' fires the simulations were never truly realistic and failed to teach firefighters how a compartment fire was likely to develop under a wide range of venting parameters.

As the training program developed, the safety procedures and simulator designs associated with Compartment Fire Behavior Training (CFBT) advanced inline with much scientific research. The intention was to produce simulators that were safe but effective in offering realistic conditions. With the basic geometry of the steel containers being ideal for creating repeatable evolutions of igniting fire gases, a universal approach evolved in the design and use of such facilities to teach various aspects of fire behavior. As an example, there are observation units for flashover; window units for backdraft; and attack units where 'door entry' and 'crew advancement' techniques are practiced. The design specifications and methods of use vary between each type and may offer local adaptations, whilst still conforming with the original Swedish model.  

There are strict controls of safety advised for the use of such units and these include -

1.  All firefighters should be fully hydrated before entering the simulators and re-hydrated at the end of training.

2. Both outer layers and undergarments of protective clothing should be of a high standard and include flash-hoods, ensuring all exposed skin is fully covered at all times. Clothing should be loose fitting, allowing an air-gap between undergarments. Damp clothing should not be worn inside the simulators.

3. There should be at least two hose-lines fitted with fog-nozzles available during the training. They should be fed by separate pumps and also supplies where possible. The interior line is managed by a maximum of 4-6 students and one instructor and the exterior line is managed by a safety officer and instructor. 

4.  Personnel are assigned specifically to operate ventilation hatch controls.

5.  There should be at least two points of exit available to firefighters inside the simulators. 

6.  The rear doors of observation simulators should remain open at all times during occupation of the facility.

7.  Simulators used to demonstrate 'backdrafts' should not be occupied by any personnel at any time during the training.

In 1991 the Fire Technology Laboratory of the Technical Research Center of Finland (VTT) carried out research into the operation and safe use of container style compartment fire simulators. Johan Mangs and Hakan Kruse reported on the results of this research in Fire International Magazine (UK) December/January 1992 p32-38. They carefully assessed the heat-flux and monitored temperatures at various locations, including those areas occupied by firefighters. They concluded that a 500mm x 500mm roof hatch was suitable and that the simulator design based upon the Swedish model is safe and effective for use and occupation by firefighters as a method of teaching fire behavior and gaseous-phase extinguishing techniques. They emphasized the intention was to avoid any progression to full flashover whilst the unit remained occupied and that maintaining control of the environment by cooling the gases in the overhead was critical to safety. They demonstrated maximum temperatures of 200 deg. C at shoulder height and up to 400 deg. C at top of helmet  for a few brief (2-3) seconds were experienced by kneeling students during repeated ignitions of the gas layers.

A further study by the University of Central Lancashire (UK) (K. Roughley) in 1999 reported maximum temperatures of 150 deg. C were experienced at the shoulders of crouching firefighters inside the observation simulators.

Firefighters operating inside CFBT simulators should not be subjected to temperatures in the EMERGENCY range on areas of the body other than at helmet tips. The more acceptable 'ordinary' range is a safer operating zone for temperatures and heat flux experienced below helmet level.

RECENT CFD RESEARCH INTO FIRE SIMULATORS IS FLAWED

There have been two recent research projects that have both attempted to use Computational Fluid Dynamics (CFD) to resolve situations of reported 'dangerous conditions' linked to excessive temperatures experienced at firefighter locations inside CFBT container style simulators. However, these research projects are seriously flawed in that CFD cannot (at this time) model firefighting water applications. The research was further prompted by two fire authorities who apparently failed to follow the original Swedish guidelines relating to safe practice in the simulators.

The first research report appeared in the May 2002 edition of Fire Prevention & Fire Engineers Journal (UK) where Nick Pope (p33-36) reported 'overly high temperatures within a flashover training simulator used by London Fire Brigade (at the Fire Service Training College - Moreton) had made the simulator 'dangerous' for use by trainee firefighters. He went on to describe how CFD was used to model conditions within the simulator and resolve the 'overly high temperatures' by increasing the ventilation hatches from one to three. What this research failed to account for was the water applications (pulsing water-fog) that are (should be) used to control the environmental conditions within the simulator, ensuring temperatures at firefighter locations do not become overly high. The report referred to temperatures at the entry point in excess of 600 deg. C but these were at ceiling level! Further still, the firefighters were reported as occupying an 'observation' unit and if this is the case, they would not enter AFTER the fire had been developing for some time (as stated) but would have occupied the compartment prior to ignition and observed the fire's development from its incipient stages through to 'flashover', whilst controlling the upper level temperatures with a pulsed application of water-fog. If the unit was an 'attack' unit then they would have entered sometime after the fire had begun, practicing door entry techniques and applying a cooling fog into the upper gas layers just prior to entry. 

The second research report appeared in the November 2002 edition of FIRE Journal (Australia) and the authors admitted their research was prompted by the original 'Pope' report in the UK. The Brammer & Wise research was initiated by the Australian Capital Territory (ACT) Fire Authority following similar reports of 'dangerous conditions' existing inside a CFBT container simulator. Again they resorted to CFD modeling to provide solutions to excessive temperatures experienced at firefighter locations and again they altered the ventilation arrangements to 'improve' conditions. However, again there is no mention of water applications or environmental control and it appears that the ACT firefighters were occupying the space without any water available to them at all as they observed a fire develop through and beyond its flashover stage!

The two reports concluded with recommendations for improving conditions within the simulators and yet failed to reference previous research in this field that had already dealt with these aspects. The reports also failed to account for any cooling effect of water on the gaseous-phase state and the likely influence  this might have for ensuring temperatures are controlled and maintained at safe levels. The fire authorities involved appear to have been using the training simulators outside of universally accepted safety guidelines, totally unaware of the design features and training objectives of the simulators in use.

Such research can be totally misleading if allowed to stand alone, unchallenged, and these reports could form the basis of future design specifications of CFBT simulators, suggesting to current users that their own units may be dangerous. This would be far from the truth where the Swedish design and user model has been followed. It is also unnecessary and ineffective and fire authorities using such simulators in future would be well advised to acknowledge the long history of past experience and scientific research that already ensures that, if followed, the Swedish model of CFBT simulations remains the safest and most effective option. They should also ensure that instructors are both trained and qualified under the original Swedish model and that local adaptations in design, training or use of the units are carefully reviewed for safety, with the original specifications and training objectives in mind.

Paul GRIMWOOD   www.firetactics.com

NOVEMBER 2002